Saw a random high school golf team rip chipotle together post match. Fuck I miss that shit. Enjoy it gents #imwashed

“Best rebounder in the history of basketball!” @POTUS x @dennisrodman 🇺🇸

🚨ℹ️⛳️ JUST IN — Phil Mickelson recently checked into rehab voluntarily to deal with ‘family trauma & addiction’, TMZ reports. Mickelson reportedly already completed the stint. The 56-year old has been absent from the public eye this year as he has navigated personal matters.

🚨 AMT-130 FOUR-YEAR DATA: Before you panic over “75% → 44%,” we need to talk about SURVIVOR BIAS. uniQure released its four-year AMT-130 data today. You're probably going to see headlines focusing on one number: At 48 months, high-dose AMT-130 showed 44% slowing of disease progression on cUHDRS compared with the updated ENROLL-HD natural-history control. That result was not statistically significant. But that is NOT the entire story. One of the most important numbers in today's release may actually be this: 53%. By four years, 53% of the matched ENROLL-HD control data were missing. And uniQure reports that the patients who stopped providing follow-up data were progressing materially faster than the patients who remained. Why does that matter? Imagine following 100 people with HD. Over time, the people progressing fastest become less likely to remain in an observational study — traveling gets harder, disability increases, participation becomes more burdensome, etc. Eventually, the people remaining in your dataset disproportionately represent those progressing more slowly. If you only look at who's left, it can make untreated HD appear to progress more slowly than it actually does. That's survivor bias — or, more specifically here, bias caused by non-random loss to follow-up. And when that group is your comparison for AMT-130, it matters enormously. Simple example: Treated patients decline 1 point. Untreated patients decline 4 points. ➡️ 75% slowing But imagine survivor bias makes the untreated group appear to decline only 2 points. The treated patients STILL decline exactly 1 point. Now the calculation says: ➡️ 50% slowing The treated patients didn't change. The comparison did. That's why saying simply, “AMT-130 dropped from 75% slowing to 44% slowing” leaves out an enormous part of the story. A few other important findings from today: 🔹 48 months — high dose cUHDRS: 44% slowing (p=0.144; not statistically significant) TFC: 61% slowing (nominal p=0.008) 🔹 When the SAME 48-month patients were compared with the previous ENROLL-HD control: cUHDRS: 53.5% slowing (nominal p=0.041) TFC: 68.3% slowing (nominal p=0.001) 🔹 Updated 36-month data — now 15 high-dose patients cUHDRS: 80% slowing TFC: 67% slowing And importantly, the four-year data continue to show a dose-response pattern, with the high-dose group declining less than the low-dose group. None of this means survivor bias automatically explains everything we're seeing at year four. We need to be careful about making that claim. But it DOES mean the 44% number cannot be interpreted in isolation. The percentage of “slowing” depends on BOTH sides of the equation: How the treated patients are doing AND how the comparison group is doing. If the faster-progressing people disappear from your natural-history dataset, your comparison group starts looking healthier — and the apparent treatment effect gets smaller. So please don't let a headline tell you that AMT-130 suddenly “stopped working.” The four-year data are more complicated than that. And for a progressive disease like Huntington's, understanding who is missing from the data can be just as important as understanding who is still there. #HuntingtonsDisease #AMT130 #TimeMatters